JarCastingaws-cloudsearch-scala
This is a Scala client for Amazon CloudSearch.
How to use
Add a following dependency into your build.sbt at first.
libraryDependencies += "jp.co.bizreach" %% "aws-cloudsearch-scala" % "0.0.2"
Then you can use aws-cloudsearch-scala in your code.
import jp.co.bizreach.cloudsearch4s.CloudSearch
import jp.co.bizreach.cloudsearch4s.CloudSearch._
val cloudsearch = CloudSearch(CloudSearchSettings(
registerUrl = "http://xxxx",
searchUrl = "http://xxxx"
))
cloudsearch4s can handle documents as Map[String, Any] or case class. If you want to handle documents as case class, you have to define a case class which is mapped to the index in the CloudSearch at first. If you use case class, camel case property names are mapped to lowercase with underscore.
case class Job(
jobTitle: String, // mapped to job_title
jobContent: String, // mapped to job_content
salary: Int // mapped to salary
)
Register
// Register single document by Map
cloudsearch.registerIndexByMap(
Map("job_title" -> "Title", "job_content" -> "Content")
) match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(id) => println(id) // String
}
// Register multiple documents by Map
cloudsearch.registerIndicesByMap(Seq(
Map("job_title" -> "Title", "job_content" -> "Content"),
Map("job_title" -> "Title", "job_content" -> "Content")
)) match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(ids) => println(ids) // Seq[String]
}
// Register single document by case class
cloudsearch.registerIndex(Job("Title", "Content")) match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(id) => println(id) // String
}
// Register multiple documents by case class
cloudsearch.registerIndices(Seq(
Job(Title, Content), Job(Title, Content)
)) match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(ids) => println(ids) // Seq[String]
}
Update
// Update single document by Map
cloudsearch.updateIndexByMap(
"091f2b7e-5b3b-4936-ae42-c560655a165f",
Map("job_title" -> "Title", "job_content" -> "Content")
) match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(id) => println(id) // String
}
// Update multiple documents by case class
cloudsearch.updateIndices(Seq(
("091f2b7e-5b3b-4936-ae42-c560655a165f", Job("Title", "Content")),
("00dd1a55-0e6d-437a-9e16-8f8b50f45a20", Job("Title", "Content"))
)) match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(ids) => println(ids) // Seq[String]
}
Delete
// Delete single document
cloudsearch.removeIndex("091f2b7e-5b3b-4936-ae42-c560655a165f") match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(id) => println(id) // String
}
// Delete multiple documents
cloudsearch.removeIndices(Seq(
"091f2b7e-5b3b-4936-ae42-c560655a165f",
"00dd1a55-0e6d-437a-9e16-8f8b50f45a20"
)) match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(ids) => println(ids) // Seq[String]
}
Search
cloudsearch.search(classOf[Job],
// Query which is assembled using Lucene's query builder API
new TermQuery(new Term("index_type", "crawl"))
// Highlight settings (Optional)
highlights = Seq(HighlightParam("job_title")),
// Facet search settings (Optional)
facets = Seq(FacetParam("employment_status"))
) match {
case Left(error) => println(error.messages) // CloudSearchError
case Right(result) => { // CloudSearchResult[Job]
...
}
}
UnitTest
To mock CloudSearch, use CloudSearchMock instead of an instance which is generated by the companion object.
val testService = new HogeService {
override protected val cloudsearch = new CloudSearchMock(Map(
"q=*:*" -> CloudSearchResult[Job](
total = 1,
hits = Seq(
CloudSearchDocument(
id = "091f2b7e-5b3b-4936-ae42-c560655a165f",
fields = Job("Title", "Content"),
highlight = Map.empty
)
),
facets = Map.empty
)
))
...
}
Give expected queries and its result to the constructor of CloudSearchMock. If unspecified query is given, test is failed. And it's possible to get history of request by CloudSearchMock#executedRequests. You can check whether expected requests are sent.
